Performance barriers of Civil Registration System in Bihar: An exploratory study

Objectives Vital statistics generated by the Civil Registration System (CRS) are essential for developing healthcare interventions at all administrative levels. Bihar had one of the lowest levels of mortality registration among India’s states. This study investigates CRS’s performance barriers from the perspective of CRS staff and community members in Bihar. Methods We conducted a primary qualitative survey in the two districts of Bihar during February-March 2020 with CRS staff (n = 15) and community members (n = 90). We purposively selected the Patna and Vaishali districts of Bihar for the survey. Thematic analysis was done to identify the pattern across the data using the Atlas-ti software. Results Most participants showed a good understanding of registration procedures and birth and death registration benefits. The perceived need for death registration is lower than birth registration. Birth registration was higher among female children than male children. We found that most participants did not report children or adult female death due to lack of financial or property-related benefits. Most participants faced challenges in reporting birth and death due to poor delivery of services at the registration centres, higher indirect opportunity cost, and demand of bribes by the CRS staff for providing certificates. We found a lack of adequate investment, shortage of dedicated staff, and limited computer and internet services at the registration centres. Conclusions Poor data on birth and death registration could lead decision-makers to target health services inappropriately. Strengthening health institutions’ linkage with the registration centres, mobile registration in far-flung areas and regular CRS staff training could increase death registration levels. An adequate awareness campaign on the benefits of birth and death registration is required to increase the reporting of vital events.

10. In line no. 556-57, you have mentioned that birth and death registration is not being done at sub-center or village level? Is it not community-based as of now? Review the lines Reply-In rural areas, birth and death occur at home are registered by Anganwadi sevika within 21 days of the event. In urban areas, birth and death occur at home is registered by registrar at municipal corporation and block offices. Institutional birth and death is registered by designated health staff at health centers within 21 days of events. Please see fig 1, page no. 5. We have revised the manuscript now.
11. You have mentioned in line 567 that all PHCs are not linked to CRS. Is this statement true? Please review. Reply-We have found many PHC are not linked to CRS due to lack of adequate logistics and lack of dedicated staff or negligence by the medical officer. A recent paper also showed registration are missed in PHC due to negligence (Kumar et al., 2021) Reviewer #2: Review: Performance barriers of Civil Registration System in Bihar: An exploratory study This is an interesting exploratory study, relying on interviews with government officials and rural residents, which documents the functioning of the civil registration system in Bihar before the pandemic. Given the importance of understanding and improving the civil registration system, such studies are urgently needed. I felt that the study should have been more in-depth. Interviews with CRS officials and with community members could have been more detailed, and the description of the overall state of the CRS in Bihar, despite the authors' attempts, is still quite incomplete. However, given the paucity of scholarship on the civil registration system in India, specially of a qualitative nature, this paper is welcome. I have some minor suggestions, which I hope would be helpful for the authors in revising their work.
1. The authors motivate their paper saying that it is important to study the civil registration system in Bihar because of its poor development indicators in general, and the fact that its CRS is estimated to record the lowest proportion of deaths and births. However, although the authors provide some hints at why Bihar's CRS performs so poorly, we do not get an explicit discussion of why precisely Bihar lacks even other northern Indian states, such as Madhya Pradesh or Rajasthan, which have higher registration coverage than Bihar. In this, the state of Bihar and Uttar Pradesh are way behind all the other states in India, and it is not clear why this is the case. To truly understand this, a comparative design is needed. Perhaps the case is that the lack of linkages between health systems and the CRS which the authors document, or the lack of computerization in Bihar explains why it lags behind other Indian states (even similarly poor ones). My suggestion to the authors is to consider this line of thinking carefully. They can see what the extant literature says about other poor Indian states (they may find looking at the annual CRS reports at the state level and the national level useful here) about CRS being so poor in Bihar. They should also note some lines of thinking for future researchers in the discussion section, as well as perhaps a limitation that they have not adequately addressed these questions. Reply-There are limited studies on CRS at state levels, particularly in Bihar. We could not locate any study examining why CRS is poor in Bihar. Although we welcome the reviewer's comment on a comparative picture of Bihar/UP with MP or Rajasthan, it is beyond the scope of the current study. Therefore, a comparative study cannot be done for the present study as it collected information of Bihar only. We may explore this research gap in future studies. We have included these lines in the limitation section of this manuscript. Please see line no. 560-561.
2. Related to this point, we also don't get a sense of the changes in the Civil Registration System before the pandemic in Bihar. Existing patterns, both from the data that has been accessed by journalists during the pandemic period (for instance see https://www.indiacovidmapping.org/reports/mortality/BiharFactsheet.pdf), as well as annual estimates before the pandemic suggest improvements in the Civil Registration System in Bihar. How did these improvements came about? If the authors have information on this, it would be extremely valuable. Similarly, what happened to Bihar's civil registration system during the pandemic? Answering this likely requires fresh data collection. If that is not feasible, then that should be noted in the discussion. If it is possible to call CRS staff on phones and ask some of these questions, then that would be quite valuable. If this is not possible, then perhaps the authors can consider this as a future research endeavor. Reply-We have included information regarding development of CRS in Bihar before pandemic. Also, we added what happen to CRS in Bihar during the pandemic in the revised manuscript. Please see line no. 452-455.& 459-466. 3. At a large number of places, I felt that the authors have policy recommendations that were not supported by the information they present. In some cases, their recommendations can actually be harmful. For instance, the authors recommend "mandatory burial or cremation permits." Instead of improving registration, I can see how this kind of procedure would cause even more problems in a place like Bihar. For instance, people will have to wait to get a death registered and could then only approach burial or cremation grounds? Instead of this, what would work well is a "notification" and "proof" system: burial grounds can notify CRS authorities about a death, and the family could have a slip which would help as a proof of death. These linkages exist in other states, especially with healthcare facilities. Public hospitals can in fact issue death certificates in many states, and private hospitals issue a hospital death certificate (or birth certificate) which can be taken to a registrar to get a death registered.
Officer or District Statistical Officer in such cases. A similar problems were also shown by a previous literature .We have included this sentence in revised manuscript. Please see line no. 433-434. 5. I was not convinced by the literature cited that "lack of birth registration is associated with school dropout …" and so on. It is quite likely that this is just selection, for instance poor children are more likely to drop out and poor children are less likely to have had their births registered. I would suggest dropping this citation. Reply-We tried to show the association of unregistered birth and school dropout. We don't intend to show absence of birth certificate is a causal factor of the school dropout. However, we have reframed the sentence now. Please see line no. 91-92 6. The authors cite the CRS report to report that 7 percent of births and 8 percent of deaths were not registered in 2019. Although this is an official estimate, it is likely an over-estimate. There are multiple reasons for this, among them the fact that the 2019 CRS deaths in this report are compared to 2018 SRS. Other reasons include the fact that the SRS may under-estimate the crude death rate. This line of thinking is explored in detail here (appendix 1: https://www.medrxiv.org/content/10.1101/2021.09.30.21264376v1). I leave it to the authors to decide how best to convey the uncertainty in the true rate of registration completion. Perhaps the authors can say "according to official estimates …". Reply-We have written "According to official statistics", line no. 95-96. 7. In figure 1, which provides the organization of the CRS, I was surprised that the Panchayat was not mentioned anywhere in the hierarchy. In figure 3, the panchayat sevak is mentioned as a registrar. In some other northern states, such as Madhya Pradesh, the bulk of rural death and birth registration happens at the Panchayat level. Is it the case that this does not happen in Bihar? In Madhya Pradesh, I have seen functioning computers in Panchayat offices which can register deaths. In which case, this may be the reason why death registration is so poor in Bihar -the lack of panchayat registration. The authors could perhaps explore this more, and correct figure 1 if necessary. If the panchayat is not a functional place to register deaths, then this is actually an important policy suggestion, to improve registration and computerization at the Panchayat level.  Figure 1 claims that an annual report at the "state level is published by the chief registrar" and also that "Annual report at India and state level is published by ORGI". This is confusing, and likely the latter sentence needs modification. Reply-Annual report at India level is published by ORGI, New Delhi. In Bihar, state level report is published by Directorate of Economic and Statistics, Department of planning and development. We have changed the fig 1 now. (Please see page no .5) 9. I thought that the bit about people not reporting infant deaths because it is considered a sin of parents in previous birth was quite revealing. The state government should consider potentially challenging this belief, and this also is a policy recommendation that flows from the findings from the paper. In addition, given the levels of awareness that the authors document, perhaps more information campaigns are also worth recommending. Reply-We have revised the manuscript in light of suggestions. Please see line no. 544-45 10. Similarly, I was surprised by the sentence that "Public facilities record all birth and death but don't update on the CTS portal until birth and death certificates are requested by family members." The authors do not present evidence for it. This is quite important, and also deserves to have highlighting from the perspective of policy recommendation that the paper generates. Reply-We observed health staff don't update registered data from registers to the CRS portal until birth and death certificates are requested by family members. A study documented a similar problem (Reference no. 38). We recommended for proper monitoring of all registration units. Please see line no. 548 11. In line 536: the authors say the CRS form is complex. Perhaps they should elaborate more? I do not agree with all the things the CRS form asks (for instance, whether the person smoked or drank alcohol is not collected by most other countries) but I didn't have the impression that it was very complex. If the authors think it is complex, I would like to see an explanation why. Otherwise, I would like this sentence to be reformulated. Reply-We have removed this sentence in the revised manuscript.
Reviewer #3: I am reviewing this paper from a non-expert perspective -I have limited knowledge of the civil registration system itself, and the specific issues faced in Bihar.
The paper clearly explores an important issue, namely barriers to civil registration in Bihar. The importance of strengthening civil registration has been highlighted by the pandemic, making this study timely. The methodology -interviews with key informants and members of the public -seems appropriate, and the paper includes a number of interesting insights about these barriers at both administrative and cultural levels.
I have a few comments for the authors to consider: 1. Quantitative data. Although this is a qualitative study, it would still be helpful to have some summary quantitative data. At least: amongst the focus group participants in Patna how many births and how many deaths were reported; and how many of these were registered. Plus the same for Vaishali. This is not to draw any quantitative conclusions, but merely to understand the overall sample.
In fact, some quantitative data is presented, but too unclearly. "A few participants (n=15) reported death at the registration centres situated at cemeteries or burial places" -this sentence needs a denominator: 15 out of how many? In sentences such as "Besides, two-fifth of the study participants reported the death of their household members at the block or municipality offices." it would be better to present the actual numbers of how many deaths were reported, and out of these how many at block or municipality offices.
The same comments apply at several other places in the text, for example: "Nearly one-sixth of the total participants were unaware of the death registration procedures and benefits associated with death registration. A higher number of participants from Patna applied for birth and death registration of their household members than participants from Vaishali. Rural and urban differentials in birth and death registration were found." In every case numbers, with denominators, would be better. "However, a subset of participants (n=10) did not receive a birth certificate for their child at all." Again, a denominator is needed. Also, not clear: were these people who applied for, but never received, the birth certicificate; or who never applied for the certificate? Reply-Dear reviewer, we have not explored categorically the number of registered births and deaths, in this study. However, we have shown how many participants registered birth and death of their family members. Besides, we have included denominator in the revised manuscript in light of your suggestions.
2. Context and history of death registration in the state. The authors mention inconsistency in death registration in Bihar -the lack of a clear trend (up to 2019 at least). Do they have any insights as to why there is so much year-to-year variation in the estimated levels of death registration? Are the SRS-based estimates of coverage reliable in their view? What are the factors which could cause a significant drop in registration from one year to the next? If there were changes in the systems in place which affected trends in registration, then this would be interesting to know. Did the interviews provide any hints? Reply. Our primary study revealed that birth and death registration was decreased during 2013 and 2014 due to majority of CRS staff are engaged in Bihar Assembly election duty. Thereafter, there has been continuous increase in birth and death registration due to introduction of digital registration using CRS portal and awareness campaign.
3. Poverty and marginalisation. I find these sentences problematic: "Lack of birth registration is associated with increased school dropout, child trafficking, and child labor [1,12]. The child mortality rate is higher among unregistered children [13]." It is very likely that the common factor behind these associations is poverty and marginalisation -the sentences are not formally wrong, but suggest the lack of birth registration could be a causal factor behind, say, child trafficking or infant mortality. It seems important to be more clear here.
This raises an important point: the authors really should discuss how factors such as caste and poverty are associated with levels of registration. Are there studies on this topic, relevant to Bihar? If there is a relevant literature, then it should be referenced and discussed at least briefly. It would be interesting to know whether there are examples of other states where poverty and marginalisation are comparable to Bihar, but civil registration is higher and more stable, and if the authors have some insight about why. If many people in Bihar perceive little benefit in reporting deaths, then why is this different in other comparable states? Are the practical barriers -loss of time and income, difficulty, etc., different in other states?
Although the sample may be too small to understand how attitudes to registration vary by caste/class there still seem to be some hints in the data (e.g. not registering infant deaths -"sin of previous birth"). It seems that the authors did not collect demographic data on caste/ occupation/ socioeconomic status, etc? Perhaps the authors should mention this as a limitation of the study. It would still be interesting if they have any insights from this work or the work of others on how these factors might affect access to registration or attitudes to registration. Reply-We have shown association in the sentence "Lack of birth registration is associated with increased school dropout, child trafficking, and child labor", not the causal factor. Moreover, we have reframed this sentence in the revised manuscript "A low birth registration is associated with increased school dropout, child trafficking, and child labor". We did not collect socio-economic data in this study. Also, comparison study with other similar state cannot be done at present due to limited literature. We have mention this sentence in limitation section of this study. Please see line no. -559-560 4. Recommendations. A number of the recommendations for how to improve vital registration are made by the authors, and many of these could be useful. But it is not really clear where responsibility lies for the failings. After statements like "the officers did not take any steps for improving the registration level in recent times, particularly death registration." could the authors indicate what the officers could have done? Also perhaps there could be more discussion of resources. A "lack of adequate investment" is mentioned in the abstract, and later the key informants "...suggested that adequate funds should be provided for the smooth functioning of the system." What is the resourcing of the CRS in Bihar? Did the interviews provide insight into where such funds should go (more staff? better pay? better training? more registration offices?) Here, again, some comparison with other states would be interesting.
Reply-We have revised the manuscript in light of the suggestions. However, there is limited study on CRS in India, so a comparative study cannot be done at this stage. We will explore this issue in a future study. We have included some suggestions on the role of authority in the recommendation section of this manuscript. Please see line no. 543-551 5. Minor points "Epidemiological purposes" are mentioned in the introduction -the authors could say more about how, in the context of the current pandemic, there has been an urgent neall-causel cause mortality data on account of weak official surveillance of COVID mortality. Any comments on uncertainties around pandemic mortality in Bihar and in comparable states would help to highlight the current importance of this study.
Reply-We have revised the manuscript in view of suggestions. Please see line no.

452-455
In Table S1, the fertility rates for Bihar and India appear to be exchanged. Reply-We have cored the Table S1 .Please see supplementary table (S1 table). identify the pattern across the data using the Atlas-ti software.   is exact and actual data, certified by registering authority, and legally acceptable [2]. A birth 65 certificate documents information such as age, place of birth, and family background [3]. Besides 66 official documentation of a child's birth, it facilitates government-aided essential services such 67 as education, health facilities, and social security [4][5][6]. CRS also provides timely and reliable   S2 Table). We also showed estimated birth and death 163 registration level by districts of Bihar (See S3 Table).  in Bihar. We also did telephonic interview with two CRS staff (Block Statistical Officers) to 188 assess the operational status of the system during the pandemic.  Table 1.  We anonymised all recorded notes and audio. We chose thematic analysis for obtaining a 224 systematic framework to code qualitative data to identify patterns across the data [32]. The 225 transcripts were transcribed in the local language (Hindi). Further, the recorded transcripts were 226 translated into English to understand the theme by the wider population. KK coded transcripts 227 and organised them by the sub-group of the participants. We used qualitative data software 228 Atlas-ti 8.0 for the thematic analysis of data. We used direct quotes for exemplary purposes. The 229 senior author (NS) reviewed typed transcripts for accuracy, completion, and plausibility. We 230 also looked for the data saturation to validate our findings.

239
Themes generated based on FGDs among community members 240 Based on the FGDs, we identified three main themes that were found to be critical barriers in   FGDs' participants (n=90), nearly fifty percent registered their child's birth to anganwadi sevika 249 and block officers. We found that a majority of study participants or their household female 250 members delivered a baby in a public hospital. However, a subset of these participants (20 out 251 of 60) reported their child's birth to anganwadi sevika, block officials, and Nagar Nigam offices 252 due to no availability of dedicated staff and shortage of registration forms at the public health 253 facilities. In addition, the requirement of identity cards was found to be barrier to birth 254 registration. According to the participants, identity proof (Aadhaar card, residential certificate, 255 and ration card) is required for identity verification before birth and death registration. In the 256 case of a child's birth registration, his or her parent's identity proof is required.     local customs such as naming ceremonies. We found that participants gave more importance to 327 the naming ceremony than the formal registration of the child's birth. Most participants (50 out of 90) said that there was lower importance of death registration than 330 birth registration. The death certificate was not required before cremation or burial. Some 331 participants (40 out of 90) perceived death registration should be done within the prescribed 332 time. A few participants (15 out of 90) also said it is not essential to report infant death whose 333 birth was not registered. A few participants (10 out of 90)) did not report infant deaths because 334 they believed that infant death is associated with sin of previous birth, and it should not be 335 disclosed to others.   allowed on submission of the affidavit, late fine and block and district level officers' approval. 378 We found most of the staff perceived timely and complete birth and death registration is crucial 379 for generating accurate vital statistics for policymaking. However, half of the total CRS staff 380 interviewed did not read the CRS manual.     We also interviewed two CRS staff over telephone after third wave of Covid (January 2022) in 452 India. They revealed that they did not achieve the birth and death registration level target due to 453 shortage of staff. Majority of staff were appointed in Covid management duty. Therefore, the 454 system failed to record birth and death on time during the pandemic.  birth certificate is required for child's school enrolment, vaccination and his social security [4,5].  In addition, differential reporting by sex will lead to insufficient quality data on sex differences   A knowledgeable and well-trained staff is essential for effective CRS [7]. This study showed 522 staff had a good understanding of birth and death registration law and registration procedures.

523
However, the lack of adequate training on the use of the CRS portal was a barrier to recording 524 vital events. Department lack dedicated staff deputed solely for CRS work. Multiple tasks 525 overburdened CRS officials led to compromise with the quality of vital statistics. The previous 526 study also documented lack of dedicated staff, lack of training on the use of computer systems 527 for recording birth and death was barriers to CRS functioning [9,31]. Refresher training is also 528 essential for updating and solving the fields' problems [7,38].   While support for those collecting this information is crucial, addressing socio-cultural beliefs 573 and incentives for the death registration of certain people (infants, women) is also critical.

574
Adequate awareness campaign is required to increase birth and death registration level in Bihar.   (CRVS) is exact and actual data, certified by registering authority, and legally acceptable [2]. A 68 birth certificate documents information such as age, place of birth, and family background [3].

69
Besides official documentation of a child's birth, it facilitates government-aided essential 70 services such as education, health facilities, and social security [4][5][6]. CRS also provides timely showed that administrative challenges, insufficient technical capacities, and unawareness are 87 significant challenges in death registration [9]. Accurate recording of death's cause is challenging level was recorded [7,9,27] to explore the contextual determinants of under registration [7,9,27].   Table). We 183 also showed estimated birth and death registration level by districts of Bihar (See S3 Table).  Officers and registrars at Nagar Nigam were invited for the interview. Fig 3 shows the hierarchy 207 of CRS staff in Bihar. We also did telephonic interview with two CRS staff (Block Statistical 208 Officers) to assess the operational status of the system during the pandemic. profile of all the study participants is shown in Table 1.       Most participants (50 out of 90) said that there was lower importance of death registration than 374 birth registration. The death certificate was not required before cremation or burial. Some   Nearly 20 participants were asked to come the next day due to irregular electric current and no 401 printing materials at the block offices. Besides, we found nearly twohree-fifths of the participants     There was no regular supervision at the block level or community level by the district or state 480 level officers. Additionally, there was no strict implementation of the RBD Act, and the officers 481 did not take any steps for improving the registration level in recent times, particularly death 482 registration. We found health institution does delayed registration; however, it is not allowed as   block-level officers revealed that adequate awareness campaign on importance of death 499 registration is needed to mandatory burial or cremation permits might increase death reporting.

500
In addition, they suggested that adequate funds should be provided for the smooth functioning 501 of the system.

503
We also interviewed two CRS staff over telephone after third wave of Covid (January 2022) in 504 India. They revealed that they did not achieve the birth and death registration level target due to 505 shortage of staff. Majority of staff were appointed in Ccovid management duty. Therefore, the 506 system failed to record birth and death on time during the pandemic. The inadequate knowledge and attitude of community members on birth and death registration 520 procedures and benefits are barriers to registration levels. Previous studies documented 521 unawareness among people is a significant challenge in birth and death registration [2,23,31]. In 522 our study area, knowledge on registration was not a barrier, yet we found a discouraging attitude, 523 particularly to death registration, contributed to incomplete CRS in the study area. Perception       According to our participants, a lower death registration among females and children is due to a 536 lack of legal necessity and no immediate financial or property-related benefits. We also observed 537 that the presence of social stigma for premature death is negatively impacting death registration  In addition, differential reporting by sex will lead to insufficient quality data on sex differences  periodic audit of death registration is required [9,31,3638,39]. In many countries, birth and death 561 registration is mandatory, and a fine is strictly imposed for not reporting the event within the 562 prescribed time [2,37].  A knowledgeable and well-trained staff is essential for effective CRS [7]. This study showed 578 staff had a good understanding of birth and death registration law and registration procedures.

579
However, the lack of adequate training on the use of the CRS portal was a barrier to recording 580 vital events. Department lack dedicated staff deputed solely for CRS work. Multiple tasks 581 overburdened CRS officials led to compromise with the quality of vital statistics. The previous 582 study also documented lack of dedicated staff, lack of training on the use of computer systems 583 for recording birth and death was barriers to CRS functioning [9,31]. Refresher training is also 584 essential for updating and solving the fields' problems [7,38]. [7].  Despite a comprehensive analysis, this study has some limitations. This study covers only two 617 districts of Bihar; therefore, generalisation could be made for the states or districts with a similar 618 development level with care. We did not show comparison study with CRS operations in other 619 states with similar level of development due to limited literature. We did not collected data by 620 castesocio-economic background (such as education, wealth, religion or castes) of study 621 participants; therefore, caste wisesuch analysis cannot be done in this study. AlsoIn addition, 622 this study is mainly focused on barriers at the operational level and not assessed the quality of